# Anti-generic decks: the 4-part test that stops "every strategy deck sounds the same" A client telling you "this sounds like every other strategy deck" is not a comment about wording. It means your deck could belong to any company. AI makes that worse because it writes the *average* deck: pillars, synergies, "drive growth", three icons per slide. This chapter gives you one test, one framework, one prompt and three before/after slides. --- ## The swap test (30 seconds) Take any slide. Replace your client's name with their biggest competitor's name. **If the slide still makes sense, it's generic.** Rewrite it. Try it on "We will drive sustainable growth through customer-centric innovation." Works for Nike, a bakery and a bank. That's the problem in one line. --- ## The framework: SNEF (Specific insight, Numbers, Enemy, Falsifiable claim) Every slide that matters (and the deck as a whole) needs all four. *All examples in this chapter use a fictional B2B software client. The numbers are illustrations of the shape, not real data.* ### S: Specific insight Something true about *this* business that the client didn't already say out loud. It usually comes from looking at their data, their customers or their competitors more closely than anyone did. - Generic: "Customers value convenience." - Specific: "62% of cancellations happen the week a customer's first delivery is late. [source: your support tickets, Jan–Jun]" **Where to find one:** support tickets, churn reasons, sales call notes, the 1-star reviews of the client *and* of competitors, the search terms on their site, the step where their funnel drops most. ### N: Numbers At least one number per slide that the client can check. Prefer their own numbers to industry stats. A number makes the claim testable and shows you did the work. - Generic: "Significant opportunity in the mid-market." - Numbered: "Mid-market is 18% of accounts and 41% of revenue. We spend 9% of sales time there." **If you don't have it, write `[ADD: metric, source]` on the slide.** A visible gap is better than an invented stat; the client will happily send you the number. ### E: Enemy A real thing you're fighting. Not a person, but a habit, a competitor's approach, a belief, a cost. Strategy is choosing against something. No enemy means no choice was made. - No enemy: "We will improve onboarding." - Enemy: "Our enemy is the 40-minute setup call. Competitors still require one. We'll make setup self-serve in under 10 minutes and win the customers who hate calls." Enemy types that work: the industry default ("everyone does annual contracts"), the status quo ("spreadsheets"), a cost ("3 hours of manual reconciliation a week"), a belief ("you need a big team for this"). ### F: Falsifiable claim A statement that could be proven wrong by a specific date. It shows you're willing to be measured, which is exactly what a client wants from a strategist. - Unfalsifiable: "This will strengthen our market position." - Falsifiable: "If we make setup self-serve, trial-to-paid goes from 11% to 15% by end of Q2. If it doesn't pass 13% by April 30, we roll back and test pricing instead." **The kill condition** ("if not X by date, we stop") is the single most un-AI thing you can put in a deck. Models almost never write one unprompted. --- ## Deck structure that passes SNEF (10 slides) | # | Slide | What it says | SNEF | |---|---|---|---| | 1 | Title = the claim | "Win the mid-market by killing the setup call" | E, F | | 2 | What we found | The specific insight, with the source | S, N | | 3 | Why it matters | The cost of the problem in money or time | N | | 4 | The enemy | What everyone (including us) does today, and why it loses | E | | 5 | The bet | One choice. What we do. | S | | 6 | What we stop doing | The trade-off. This slide is what makes it strategy. | E | | 7 | How it works | The mechanism in 3 steps | S | | 8 | Targets and kill condition | Numbers, dates, the "if not, we stop" | N, F | | 9 | First 30 days | Named owners, dated actions | N | | 10 | The ask | Exactly what you need approved today | | Slide titles are **full sentences that state the point** ("Mid-market is 41% of revenue and 9% of our sales time"), not labels ("Market Overview"). A reader who only reads the titles should get the whole argument. --- ## The prompt (any model) ``` You are a strategy partner who hates generic decks. Help me build a deck that passes the SNEF test: S = a specific insight about THIS business that the client hasn't said out loud N = at least one checkable number per slide, preferably the client's own data E = a real enemy (an industry default, a status quo, a cost or a belief) we choose to fight F = a falsifiable claim with a target, a date and a kill condition ("if not X by DATE, we stop and do Y") Rules: - Never invent numbers, sources, quotes or company facts. Where a number is needed, write [ADD: metric, source]. - Slide titles are full sentences stating the point. Someone reading only titles should get the argument. - Include a "What we stop doing" slide. - Banned: pillars, synergy, drive growth/value, holistic, best-in-class, customer-centric, data-driven, actionable insights, north star, leverage, unlock, move the needle, paradigm shift, "strategic priorities" with 3 icons. - Apply the swap test to every slide: if the client's name could be replaced by a competitor's and the slide still works, rewrite it. Context: - Client / business: [WHO, WHAT THEY SELL, TO WHOM] - The decision this deck must get: [e.g. "approve a 2-month onboarding project, budget $X"] - Data I have: [PASTE NUMBERS, CHURN REASONS, SUPPORT TICKET THEMES, FUNNEL DATA, CUSTOMER QUOTES] - What competitors do: [WHAT YOU KNOW] - Audience in the room: [ROLES] First, give me 3 candidate SPECIFIC INSIGHTS from my data and the ENEMY each implies. Ask me to pick one. Then write the 10-slide outline: for each slide, title sentence + 2–4 lines of body + which SNEF parts it covers + any [ADD] gaps. End with a swap-test report: list any slide that would still work for a competitor. ``` **Deck already written?** Paste it and use: *"Run the swap test and SNEF test on every slide. Table: slide · passes swap test? · S/N/E/F missing · rewrite of the title sentence. Then list the [ADD] numbers I need to collect."* --- ## 3 before/after slides Same fictional client: a B2B project-management tool with a sales-assisted trial. ### Slide 1: The "strategic pillars" slide **Before** > **Our Strategic Pillars** > 🚀 **Innovation**: Driving cutting-edge solutions > 🤝 **Customer-Centricity**: Putting customers at the heart of everything > 📈 **Scalability**: Building for sustainable growth Swap test: fails (works for any company). S: no. N: no. E: no. F: no. **After** > **Title:** We grow next year by fixing setup, not by shipping features. > **Body:** > - Accounts that finish setup in week 1 renew at [ADD: %]. Those that don't: [ADD: %]. (Source: renewals 2025, CRM) > - Today setup needs a 40-minute call. 3 in 10 trials never book it. [ADD: confirm from calendar data] > - So for 2 quarters, the product team works on self-serve setup only. Feature requests wait. > **Footer:** If self-serve setup doesn't raise week-1 completion to 70% by June 30, we go back to the roadmap. Covers: S (setup predicts renewal), N (the rates), E (the 40-minute call, feature-chasing), F (70% by June 30, with a rollback). ### Slide 2: The market opportunity slide **Before** > **Market Opportunity** > The global project management software market is experiencing significant growth, driven by the rise of remote work and digital transformation. This presents a tremendous opportunity for our platform. Swap test: fails. Also uses an unsourced market claim. **After** > **Title:** Our best customers are architects, and we've never marketed to them. > **Body:** > - Architecture firms are [ADD: X]% of accounts but [ADD: Y]% of revenue and churn least. (Source: billing export, Sept) > - They found us through 2 YouTube reviews, not our ads. (Source: signup survey, "how did you hear about us", [ADD: n] answers) > - Not one of our 4 main competitors has an architecture page. > **Footer:** Test: one landing page + 10 firms contacted directly. If we don't get 15 architect trials in 6 weeks, we drop the segment. Covers: S (a hidden segment), N, E (competitors ignoring it, our own generic marketing), F (15 trials in 6 weeks). ### Slide 3: The recommendation slide **Before** > **Recommendations** > - Leverage data-driven insights to optimize the customer journey > - Enhance cross-functional collaboration > - Explore new growth opportunities in emerging markets Swap test: fails. No owner, no date, no number. **After** > **Title:** Three changes, three owners, one review date: March 15. > > | Change | Owner | Target by Mar 15 | If we miss it | > |---|---|---|---| > | Remove credit card from trial signup | [ADD: name], Growth | Trial starts +30% | Put the card back, test a shorter trial | > | Reply to every churn survey within 24h | [ADD: name], Support | 20 saved accounts | Stop; move effort to onboarding | > | Annual plan at 2 months free | [ADD: name], Finance | 15% of new MRR on annual | Revert to monthly-only | > > **We are not doing:** a rebrand, a new market, a mobile app. Not this half. Covers: S (concrete changes), N (targets), E (the "not doing" list), F (every row has a miss condition). --- ## Limits - SNEF won't produce an insight from nothing. If you have no data, interviews or tickets, the honest first slide is "What we need to find out in 2 weeks." - Falsifiable claims can be wrong. That's the point, but agree with the client in advance that a missed target is information, not a failure. - Industry-wide stats can still appear, but only with a named, dated source, and never as the main insight. Made by Dan Shipped. Follow @danshipped for the next drop.